-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathRandomForestClass.py
203 lines (176 loc) · 6.59 KB
/
RandomForestClass.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
# encoding = utf8
'''
@Author: King
@Date: 2019.05.16
@Purpose: 机器学习算法学习与实现
@Introduction: A机器学习算法学习与实现
@Datasets:
@Link :
@Reference :
'''
import pandas as pd
import numpy as np
from sklearn.model_selection import train_test_split
from algorithm.ML.RandomForest.xiechengRF.MyLog import MyLog
from sklearn.ensemble import RandomForestClassifier
from algorithm.ML.matplotlibTest.drawLine import drawLineClass
class RandomForestClass:
#初始化函数
def __init__(self,data,test_size=0.3,random_state=300,n_estimators=10,max_depth=5):
self.data=data
self.data=self.pretreatment() #缺失值处理
#self.print()
self.X=[]
self.y=[]
self.X_train=[]
self.X_test=[]
self.y_train=[]
self.y_test=[]
self.test_size=test_size #指定参数test_size=0.3,数据样本作为测试集的比例为30%，输出训练集和测试集大小
self.random_state=random_state #随机发生器If int, random_state is the seed used by the random number generator;
# If RandomState instance, random_state is the random number generator;
# If None, the random number generator is the RandomState instance used by np.random.
self.divided()
self.y_pred=[]
self.model=self.RandomForestMain(n_estimators=n_estimators,max_depth=max_depth) #其构建随机森林分类模型，指定n_estimators参数为10，
# 即使用10棵决策树构建模型。将训练集传入模型进行模型训练。
self.predict()
self.MAPE=self.calMAPE()
self.CV=self.calCV()
self.RMSE=self.calRMSE()
self.MAD=self.calMAD()
#1 预处理
def pretreatment(self):
mylog.info("run pretreatment function")
data = self.data.fillna(self.data.mean())
#self.print(data=data,type=0)
return data
#2 划分数据集
def divided(self):
mylog.info("run divided function")
self.X=self.data.iloc[:,:7]
self.y = self.data.iloc[:,7]
#self.print(type=0,data=self.X)
#self.print(type=0, data=self.y)
#print("self.test_size",self.test_size)
#print("self.random_state", self.random_state)
self.X_train, self.X_test, self.y_train, self.y_test = train_test_split(self.X, self.y, test_size=self.test_size, random_state=self.random_state)
#3 构建随机森林模型并训练
def RandomForestMain(self,n_estimators=10,max_depth=5):
mylog.info("run RandomForestMain function")
model = RandomForestClassifier(n_estimators=n_estimators,max_depth=max_depth)
model.fit(self.X_train, self.y_train)
return model
#4 利用随机森林模型预测分类
def predict(self):
mylog.info("run predict function")
self.y_pred = self.model.predict(self.X_test)
#print("Predictions of test set:\n%s" % self.y_pred)
#计算平均百分比偏误差 MAPD
def calMAPE(self):
mylog.info("run calMAPE function")
i = 0
MAPEerror = 0
for y_test_item in self.y_test:
MAPEerror = MAPEerror + (abs(y_test_item - self.y_pred[i]) / y_test_item)
i = i + 1
MAPE = (MAPEerror / len(self.y_test)) ** (1 / 2)
print("MAPE:",MAPE)
return MAPE
#计算变异系数 CV
def calCV(self):
mylog.info("run calCV function")
i = 0
CVerrors = 0
for y_test_item in self.y_test:
CVerrors = CVerrors + (y_test_item - self.y_pred[i]) ** 2
i = i + 1
CV = (CVerrors / np.mean(self.y_test))
print("CV:", CV)
return CV
#计算均方根误差 RMSE
def calRMSE(self):
mylog.info("run calRMSE function")
i = 0
RMSEerrors = 0
for y_test_item in self.y_test:
RMSEerrors = RMSEerrors + (y_test_item - self.y_pred[i]) ** 2
i = i + 1
RMSE = (RMSEerrors / len(self.y_test)) ** (1 / 2)
print("RMSE:", RMSE)
return RMSE
#计算平均绝对偏差 MAD
def calMAD(self):
mylog.info("run calMAD function")
i = 0
MADerror = 0
for y_test_item in self.y_test:
MADerror = MADerror + abs(self.y_pred[i] - y_test_item)
i = i + 1
MAD = MADerror / len(self.y_test)
print("MAD:", MAD)
return MAD
#打印变量
def print(self,data="",type=1):
if type ==1:
print(self.data)
else:
print(data)
if __name__ == '__main__':
data = pd.read_csv("normalizedData.csv", index_col=0)
mylog = MyLog()
mylog.info("Start")
MAPEarr=[]
CVarr=[]
RMSEarr=[]
MADarr=[]
for i in range(1,16):
rf = RandomForestClass(data,0.3,300,i,4)
MAPEarr.append(rf.calMAPE())
CVarr.append(rf.calCV())
RMSEarr.append(rf.calRMSE())
MADarr.append(rf.calMAD())
print("MAPEarr:",MAPEarr)
print("CVarr:", CVarr)
print("RMSEarr:", RMSEarr)
print("MADarr:", MADarr)
x=range(1,len(MAPEarr)+1)
drawLine=drawLineClass(0)
#画 平均百分比偏误差 MAPD
drawLine.fig, drawLine.ax=drawLine.createSubplots()
drawLine.setX(x)
drawLine.setXLabel("树深度")
drawLine.setYLabel("MAPE")
drawLine.setY(MAPEarr)
drawLine.setLineColor("green")
drawLine.setlegendLabel("平均百分比偏误差 MAPD")
drawLine.setLineType("-.")
drawLine.drawLine()
# 画 计算变异系数 CV
drawLine.fig, drawLine.ax =drawLine.createSubplots()
drawLine.setY(CVarr)
drawLine.setXLabel("树深度")
drawLine.setYLabel("CV")
drawLine.setLineColor("red")
drawLine.setlegendLabel("计算变异系数 CV")
drawLine.setLineType("--")
drawLine.drawLine()
# 画 计算均方根误差 RMSE
drawLine.fig, drawLine.ax =drawLine.createSubplots()
drawLine.setY(RMSEarr)
drawLine.setXLabel("树深度")
drawLine.setYLabel("RMSE")
drawLine.setLineColor("yellow")
drawLine.setlegendLabel("计算均方根误差 RMSE")
drawLine.setLineType("*-")
drawLine.drawLine()
# 画 计算平均绝对偏差 MAD
drawLine.fig, drawLine.ax =drawLine.createSubplots()
drawLine.setY(MADarr)
drawLine.setXLabel("树深度")
drawLine.setYLabel("MAD")
drawLine.setLineColor("orange")
drawLine.setlegendLabel("计算平均绝对偏差 MAD")
drawLine.setLineType("+-")
drawLine.drawLine()
drawLine.show()